Restarting a job with different k points

If a calculation needs large number of k-points, such as electrode calculation in transport, I have a memory shortage problem.

To solve this, I converged the DM with different k-points gradually. For example, I need kz = 100. In this case, I tried to start calculation with smaller k points (e.g. kz = 10). After getting converged DM with kz = 10, I increased the k-points to 20. By repeating this, at the end, I got the converged DM with kz = 100.

Is there any problems or issues in this strategy?